Extending LIME for Business Process Automation
This addresses the need for explainable AI in high-stakes business decisions, but it is incremental as it builds on LIME.
The paper tackled the problem of explaining AI decisions in business process automation, where existing methods like LIME fail due to task ordering and constraints, and proposed an extension that showed advantages in empirical evaluation.
AI business process applications automate high-stakes business decisions where there is an increasing demand to justify or explain the rationale behind algorithmic decisions. Business process applications have ordering or constraints on tasks and feature values that cause lightweight, model-agnostic, existing explanation methods like LIME to fail. In response, we propose a local explanation framework extending LIME for explaining AI business process applications. Empirical evaluation of our extension underscores the advantage of our approach in the business process setting.